A working model to decide tier prices with your team. Pick the AI model you'll run, then edit any figure — prices, caps, cost assumptions. Margins recompute live. The AI-cost line is the worst case: a customer consuming their entire included token allowance (the hard-stop cap).
Blended cost / 1M tokens
—
Cost assumptions — tune to your workload; the model select fills input/output rates
$
$
%
%
$
$
%
Top-up pack — extra tokens + conversations, bought mid-cycle